A non-patent document 1 listed below proposes a modular multilevel converter (MMC) by using a switching device (for example, IGBT: Insulated-gate bipolar transistor) capable of ON/OFF control, as a method of a power conversion device which can output a high voltage higher than a breakdown voltage of the switching device.
First, a name of each part of MMC is defined in order to explain a circuit configuration of MMC.
In MMC, a bidirectional chopper circuit shown in FIG. 4 forms a unit converter.
Each unit converter is connected to an external circuit through at least two terminals. In the present embodiment, the two terminals are called an x-terminal and a y-terminal, respectively. In addition, a voltage of the x-terminal against a standard voltage of the y-terminal is called a cell voltage.
If a voltage of an energy storage device 405 of the bidirectional chopper circuit shown in FIG. 4 is denoted by VC, an available value of the cell voltage is two, VC and zero.
In the embodiment, a circuit that cascade-connects the x-terminal and the y-terminal of one or a plurality of the unit converters is called a converter arm.
Each converter arm has at least two terminals. In the embodiment, the two terminals are called a a-terminal and a b-terminal, respectively. In addition, a voltage of the a-terminal against a standard voltage of the b-terminal is called an arm voltage. The arm voltage is a sum of cell voltages of unit converters included in the converter arm.
Since the arm voltage is the sum of cell voltages, the arm voltage becomes multiples of a voltage VC of an energy storage device provided in each cell.
In the embodiment, a circuit where one terminal of a first reactor is connected in series to the b-terminal of a first converter arm, one terminal of a second reactor is connected in series to the other terminal of the first reactor, and the a-terminal of a second converter arm is connected in series to the other terminal of the second reactor, is called a leg.
The a-terminal of the first converter arm is called a P-terminal, a connection point of the two reactors is called a M-terminal of the leg, and the b-terminal of the second converter arm is called a N-terminal of the leg. Therefore, each leg has at least three terminals that are the P-terminal, M-terminal and the N-terminal. In addition, in the embodiment, a sum of arm voltages of the two converter arms included in the leg is called a leg voltage.
Since the leg voltage is a sum of arm voltages, the leg voltage also becomes multiples of the voltage VC of the energy storage device provided in each cell.
Next, an explanation will be given of a circuit configuration of MMC. Here, as an example, a three-phase MMC will be described.
The P-terminals of three legs are connected to each other and one terminal is drawn out from the connection point, similarly, the N-terminals of the three legs are connected to each other and the other terminal is drawn out from the connection point, then, a three-phase MMC can be configured. In the embodiment, the drawn out terminal from the three P-terminals connected to each other is called a positive output terminal of MMC and the drawn out terminal from the three N-terminals connected to each other is called a negative output terminal of MMC.
A DC load can be connected between the positive output terminal and the negative output terminal of MMC.
A three-phase power system can be connected to three M-terminals of the three legs. In the embodiment, the three M-terminals of the three legs are generally called a three-phase terminal.
Hereinafter, a brief explanation will be given of an operation of MMC. It is assumed that the three-phase terminal is interconnected with a three-phase power system through a transformer or an interconnection reactor.
Voltages among three-phase terminals can be controlled by controlling an arm voltage of each converter arm configuring the MMC.
For example, if a system frequency component of a voltage among the three-phase terminals is controlled to be identical to a frequency and amplitude of a system line voltage and only a phase thereof is slightly delayed in comparison with that of the system line voltage, an active power flows into the three-phase MMC from the power system.
In addition, if a system frequency component of a voltage among the three-phase terminals is controlled to be identical to a frequency and amplitude of a system line voltage and only a phase thereof is slightly advanced in comparison with that of the system line voltage, an active power flows into the power system from the three-phase MMC.
If a system frequency component of a voltage among the three-phase terminals is controlled to be identical to a frequency and phase of a system line voltage and only an amplitude thereof is slightly increased in comparison with that of the system line voltage, an advanced reactive power is generated between the three-phase MMC and the power system.
If a system frequency component of a voltage among the three-phase terminals is controlled to be identical to a frequency and phase of a system line voltage and only an amplitude thereof is slightly decreased in comparison with that of the system line voltage, a delayed reactive power is generated between the three-phase MMC and the power system.
However, there are two problems in MMC described later.
The first problem is that a reactor is required for each leg.
In the three-phase MMC, if an energy storage device included in each unit converter is an ideal DC voltage source and if voltages of all ideal DC voltage sources are equal to each other, leg voltages of three legs can be matched by properly controlling a switching timing of each unit converter.
However, in the actual MMC, an electrolytic capacitor or a battery is used as the energy storage device.
Since each unit converter operates as a single phase converter, an instantaneous power flowing into/out from the each unit converter has a double frequency pulsating component of a frequency of power system connected to the three-phase terminals or of a frequency of a three-phase load.
In addition, since MMC transmits and receives a DC power to and from a DC load which is connected between a positive output terminal and a negative output terminal, an instantaneous power flowing into/out from each unit converter also has a pulsating power component accompanying the transmission and reception of the electric power to and from the DC load.
Therefore, a voltage between both ends of an energy storage device (for example, an electrolytic capacitor or battery) included in each unit converter is pulsating, and an instantaneous value of the pulsating component is different for each leg.
As described above, a leg voltage is multiples of the voltage VC of the energy storage device included in the leg.
When the voltage VC of the energy storage device is different for each leg, it is impossible to always and completely match leg voltages of three legs even if a switching timing of the unit converter is properly controlled.
During a period that the leg voltages of the three legs are not matched, a difference among the leg voltages is absorbed by only two reactors included in each leg.
When an inductance of the two reactors included in each leg is zero, the difference among the leg voltages is absorbed by only a wiring that connects among legs during the period. Since an impedance of the wiring is small, an overcurrent flows into the three legs.
Therefore, a reactor is essential for each leg in order to control the overcurrent.
The second problem is that the reactor becomes large when MMC transmits a DC power to a DC load.
When MMC transmits a DC power to a DC load, it is required to apply a zero-phase DC current to a converter arm of each phase and a reactor. Therefore, a current that the zero-phase DC current is superimposed on a normal-phase/reverse-phase current flows in the reactor.
In this case, a maximum current value becomes large in comparison with the case that only the normal-phase/reverse-phase current flows in the reactor. In order to prevent a magnetic saturation even in the maximum current value, an increase of iron core cross section of the reactor is required, and thereby the reactor becomes large.
In addition, a non-patent document 2 discloses an illustration of DC transmission system shown in FIG. 35 with one line.
The DC transmission system of FIG. 35 includes three-phase AC power systems 3100, 3170, a breaker 3124 that is disposed in order to disconnect a DC transmission system 3800 from the three-phase AC power systems 3100, 3170, a conversion transformer 3805 that transforms an AC voltage, a three-phase full-bridge power conversion device 3801 that uses a plurality of semiconductor switching devices, and capacitors 3802, 3803 that are connected in parallel with the three-phase full-bridge power conversion device 3801, and a neutral point 3806 that is connected to the capacitors 3802, 3803 is grounded. In addition, the DC transmission system includes a DC transmission cable 3807 and a DC reactor 3804 that is connected to the power conversion device 3801 and the DC transmission cable 3807 in series.
Generally, a DC transmission system transmits electric power from a three-phase AC power system to another three-phase AC power system.